SAMUELSON_F&E ARTICLE (DO NOT DELETE) 1/31/2017 12:08 PM FUNCTIONALITY AND EXPRESSION IN COMPUTER PROGRAMS: REFINING THE TESTS FOR SOFTWARE COPYRIGHT INFRINGEMENT Pamela Samuelson* Abstract Courts have struggled for decades to develop a test for judging infringement claims in software copyright cases that distinguishes between program expression that copyright law protects and program functionality for which copyright protection is unavailable. The case law thus far has adopted four main approaches to judging copyright infringement claims in software cases. One, now mostly discredited, test would treat all structure, sequence, and organization (SSO) of programs as protectable expression unless there is only one way to perform a program function. A second, now widely applied, three-step test calls for creation of a hierarchy of abstractions for an allegedly infringed program, filtration of unprotectable elements, and comparison of the protectable expression of the allegedly infringed program with the expression in the second program that is the basis of the infringement claim. A third approach has focused on whether the allegedly infringing elements are program processes or methods of operation that lie outside the scope of protection available from copyright law. A fourth approach has concentrated on whether the allegedly infringing elements of a program are instances in which ideas or functions have merged with program expression. This Article offers both praise and criticism of the approaches taken thus far to judging software copyright infringement, and it proposes an alternative unified test for infringement that is consistent with traditional principles of copyright law and that will promote healthy competition and ongoing innovation in the software industry. I. INTRODUCTION The paradigmatic roles of copyright and patent laws have been, respectively, to protect original authorial expressions from illicit copying, and to protect novel and nonobvious functional designs (if * Richard M. Sherman Distinguished Professor of Law, Berkeley Law School. I am very grateful to Kathryn Hashimoto for excellent research for and editing of this article. I am also grateful to Clark Asay, Jonathan Band, Joshua Bloch, Oren Bracha, Dan Burk, Julie Cohen, Joe Craig, Charles Duan, Shubha Ghosh, Ariel Katz, Peter Lee, Mark Lemley, Glynn Lunney, Corynne McSherry, Christina Mulligan, Aaron Perzanowski, Michael Risch, Christopher Jon Sprigman, Fred von Lohmann, and Phil Weiser for comments on an earlier draft of this article. I also wish to thank Lionel Bently for the opportunity to give the 10th Annual International IP Lecture at Emanuel College at Cambridge University on which this Article was initially based. Electronic copy available at: https://ssrn.com/abstract=2909152 SAMUELSON_F&E ARTICLE (DO NOT DELETE) 1/31/2017 12:08 PM 2 they have been appropriately claimed and examined by patent officials) from illicit uses.1 It would be convenient if copyright law could be assigned the role of protecting the expression in computer programs and patent law the role of protecting program functionality. While courts continue to try to distinguish between program expression and program functionality, this distinction has proven elusive in the decades since the U.S. Congress decided to extend copyright protection to computer programs.2 Leading up to the enactment of the Copyright Act of 1976,3 members of Congress were warned that the functionality of computer programs would make it difficult to fit them into the copyright realm.4 But lingering concerns about the potential misfit were for a time allayed by a 1978 National Commission on New Technological Uses of Copyrighted Works (CONTU) report that endorsed copyright protection for programs.5 CONTU observed that “the distinction between copyrightable computer programs and uncopyrightable processes or methods of operation does not always seem to ‘shimmer with clarity,’” but it was nevertheless “important that the distinction between programs and processes be made clear.”6 The report expressed optimism that traditional principles of 1 See J.H. Reichman, Legal Hybrids Between the Patent and Copyright Paradigms, 94 COLUM. L. REV. 2432, 2448–53 (1994) (describing the classical patent and copyright paradigms in the international intellectual property system). 2 While some claim that Congress extended copyright protection to computer programs when it enacted the Copyright Act of 1976, there is some ambiguity in the legislative history on this point. Compare FINAL REPORT OF THE NAT’L COMM’N ON NEW TECHNOLOGICAL USES OF COPYRIGHTED WORKS 15–16 (1978) [hereinafter CONTU REPORT] (concluding that Congress had extended copyright protection to software in 1976) with Pamela Samuelson, CONTU Revisited: The Case Against Copyright Protection for Computer Programs in Machine-Readable Form, 1984 DUKE L.J. 663, 694-96 (1984) [hereinafter Samuelson, CONTU Revisited] (suggesting that §117 preserved the status quo of unprotectability under the prior act as to computer-related subject matters). However, any such ambiguity was resolved by 1980 amendments to the Copyright Act of 1976 (1976 Act), which implemented legislative changes that CONTU recommended in its report. See Pub. L. No. 96-517, 94 Stat. 3015 (codified at 17 U.S.C. §§ 101, 117 (1980)). 3 Pub. L. No. 94-553, 90 Stat. 2541 (1976), codified at 17 U.S.C. § 101, et seq. 4 See Hearings Before the Subcomm. on Patents, Trademarks, and Copyrights of the S. Comm. on the Judiciary Pursuant to S. Res. 37 on S. 597, 90th Cong. 192-97 (1967), reprinted in 9 OMNIBUS COPYRIGHT REVISION LEGISLATIVE HISTORY 192-97 (George S. Grossman, ed., 1976) (testimony of Professor Arthur Miller). Miller expressed concern that courts might construe copyright protection for programs as “extend[ing] to or embody[ing] the process, scheme, or plan that the program uses to achieve a functional goal,” saying this would confer “patent like protection under the guise of copyright.” Id. at 197. Congress responded to these concerns by adopting a provision stating that “[i]n no case does copyright protection for an original work of authorship extend to any . procedure, process, system [or] method of operation . regardless of the form in which it is . embodied in such work.” 17 U.S.C. § 102(b). This provision is discussed at length infra Part III. 5 CONTU REPORT, supra note 2, at 1–2. CONTU acknowledged that there was not “universal agreement” about copyright protection for software. Id. at 20–21. See Stephen Breyer, The Uneasy Case for Copyright: A Study of Copyright in Books, Photocopies, and Computer Programs, 84 HARV. L. REV. 281, 344–46 (1970) (questioning the economic case for extending copyright protection to computer programs). While CONTU was deliberating about new technology issues, the World Intellectual Property Organization was considering a sui generis form of intellectual property protection for software. See WORLD INTELLECTUAL PROP. ORG., INT’L BUREAU, MODEL PROVISIONS ON THE PROTECTION OF COMPUTER SOFTWARE (1978). Whether computer programs should be protected by copyright law was not part of CONTU’s original charter, which mainly focused on photocopying and digitizing published texts. See Samuelson, CONTU Revisited, supra note 2, at 663 n.2, 699. That may explain why none of the Commissioners had any expertise about computers or computer programs. Id. at 699. 6 CONTU REPORT, supra note 2, at 18. Electronic copy available at: https://ssrn.com/abstract=2909152 SAMUELSON_F&E ARTICLE (DO NOT DELETE) 1/31/2017 12:08 PM 3 copyright law, when applied to programs, would strike the right balance,7 and it was content to leave the difficult (and perhaps “futile”) task of drawing boundaries between program expression and functionality to the judiciary.8 Unfortunately, CONTU failed to fully understand the intrinsic functionality of computer programs, the importance of standards and network effects in the software industry, and the need of programmers to develop software capable of interoperating with other programs. It also failed to offer guidance on how, when, and why functionality should constrain the scope of copyright protection in programs.9 Commentators have debated for decades how much legal protection software developers should get from copyright law to induce optimal levels of investment in the development of computer programs.10 Some have worried that copyright protection for programs might either be too “weak” if infringement could be easily avoided by rewriting the same program in different source code, or too “strong” if programmers felt compelled to do things differently than an existing program in order to avoid infringement, thereby impeding beneficial standardization.11 That concern has manifested itself in the software copyright cases that followed. Appellate courts have taken four main approaches to distinguishing the original expression in computer programs from program functionality. A first-in-time, but now much discredited, approach was adopted by the Third Circuit Court of Appeals in Whelan Associates v. Jaslow Dental Lab., Inc., under which the “structure, sequence, and organization” (SSO) of computer programs was 7 See id. at 12–23. CONTU thought copyright should grant no more economic power than was needed to create proper incentives to create software. Id. at 12. 8 Id. at 22–23. 9 According to CONTU, programs were no more functional than sound recordings, id. at 10, which was simply not true. After all, the inherent purpose of computer programs is to automate functional